aggregate_redis_keys.py

import sys
from collections import defaultdict

# To be used like this:
# redis-cli --scan | python aggregate_redis_keys.py

prefixes = (
	'dev_',
	'prod_',
	'staging_',
)

categories = (
	'foo_views_',
	'bar_views_',
	'foobars_',
)

prefix_counts = defaultdict(int)
category_counts = defaultdict(lambda: defaultdict(int))
total_count = 0
for line in sys.stdin:
	line = line.rstrip()

	longest_prefix = ''
	for prefix in prefixes:
		if line.startswith(prefix):
			prefix_counts[prefix] += 1
			if len(prefix) > len(longest_prefix):
				longest_prefix = prefix

	line = line[len(longest_prefix):]

	found = False
	for category in categories:
		if line.startswith(category):
			category_counts[longest_prefix][category] += 1
			found = True
			break

	if not found:
		print('Did not found %s in list of categories' % line)

	total_count += 1

for prefix, counts in category_counts.iteritems():
	for category, count in sorted(counts.iteritems(), key=lambda x: x[1]):
		print("% 10d (% 2.f%%)        [% 30s]    %s" % (count, count * 100 / total_count, prefix, category))